Need some help here, cant seem to get midi to lock dead on =

interloper · Post by **interloper** » Fri Sep 05, 2003 11:28 am

I would not run my percussion live (meaning, from drumsynths), which is why I suggested earlier that you bring it into the audio realm, in addition to the other approaches.

Cw sampler latency is very low, but there again you are using samples and not a live instrument like kick me, therefore it should be more accurate.

I do think that if you change the way you work on your percussion you can bypass this anomaly. I used battery too, in the past, and just bounced the audio files down.

I have noticed that even with some synth bass patterns, that there is a drift, perhaps not in timing, but in the envelope stage. This created a timing jitter, or phasing, and was correctable by using one slice. For electronic music, that's the way to go.

Basic Pitch · Post by **Basic Pitch** » Fri Sep 05, 2003 12:53 pm

Inter,

For me personally the way I always worked in the past due to CPU headroom, was to create my sample accurate drum parts in battery, then render them to audio, the same as you suggested that was I could then free up CPU speed etc top run more parts.

Now I suppose I could write a part then render it to audio, so I get that 1 bar best take method, but then I am limited and if I decide I wanna change something down the road I may/may not be SOL

Then you also come into play with, if I am going to bounce everything to audio then use effects realtime, did I need a 15 DSP card to accomplish this lego building process in the 1st place, this could easily be done with a 6 DSP pulsar then no? I mean just write synth part and apply effects at mix down..

Cheers

alfonso · Post by **alfonso** » Fri Sep 05, 2003 4:14 pm

When you'll put SFP fx here and there, synths and modular stuff with polypohny, you will appreciate all your dsp's...I have 25 and I'm happy I can have top notch oscillators and filters with many voices, Vocoder, Vincos, Masterverbs and a lot of other wonderful stuff where my own sound design makes my music different from another one.....native's destiny is clichè, SFP is customization and personality, IMO...

cheers.

astroman · Post by **astroman** » Fri Sep 05, 2003 6:02 pm

On 2003-09-05 13:53, Basic Pitch wrote:
...did I need a 15 DSP card to accomplish this lego building process in the 1st place, this could easily be done with a 6 DSP pulsar then no? ...

Basic Pitch though that statement makes sense at your current perspective, I agree with Alfonso - get as much DSPs as you can afford.
You're just starting to explore the system and even your prefered way of working might change over time. As well as devices may change or a solution for your current probs shows up.

At least the developers at Adern are very aware of any possible degradation in sound quality due to 'dirty' processing and even lock modules to certain DSPs (normally SFP allocates this dynamically) to guarantee phase accuracy.

While Pulsar/Scope is on the market for some time now, that approach to precision (and the awareness thereof) is rather new.

As I've already mentioned the latest Wavelenght synth just delivers a single voice on my 4 DSPs - and I'd really like to have a second instance and feed the beast back to itself

That rather simple setup would end at the use of 8 DSPs with a poly of only four voices on each instance - but I'm shure it would crash windows or open the door to another dimension

cheers, Tom

<font size=-1>[ This Message was edited by: astroman on 2003-09-05 21:19 ]</font>

sinix · Post by **sinix** » Fri Sep 05, 2003 7:15 pm

Believe me, I render enough stuff to audio as it is and these new sequencer "freeze" options will help even more, but I totally agree with the others here.

I really wish I had more than 6 DSP's! I always run my sustaining instruments like pads live and a Pulsar II gets used up *real* quick with a decent reverb, Vinco, a delay, and a good synth.

Cheers

interloper · Post by **interloper** » Sun Sep 07, 2003 8:38 pm

With 27 DSP's I still bounce my percussion parts down. Since I'm recording at 32 bit, I don't loose anything if I process a hi-hat audio file with the Vinco again.

I understand the temptation of wanting to run everything live, but committing to audio will help your approach. It will increase your timing accuracy for parts and if you stay in 32 bit, you can further mangle the parts. Running live might not be the best approach at this time, but you should be the judge. You're the one spending time on your machine, try it both ways, and pick whatever suits you best.